The scent of a rose fades over time, and has for hundreds of years.
For centuries, generations of breeding in the quest for longer blooms and petals in shades of nearly every hue have dulled the sweetest smells that once perfumed gardens around the world.
French researchers have now figured out precisely which genes make a rose smell so sweet, and where to tinker in the genome to enhance its distinctive scent.
Although the rose genome has been mapped before, a newly published version is far more complete, indicating which genes tend to travel together — scent and color, for instance — and which genes are responsible for continuous blooming, among other traits.
“I think it’s a huge improvement on the current rose sequence,” said Rob Martienssen, a plant biologist and professor at Cold Spring Harbor Laboratory on Long Island.
“A lot of these genes were known before, but it’s a very nice way of putting them all together and showing their history. And I think it’ll be very important for breeding,” said Dr. Martienssen, who was not involved in the new study.
The new sequence is one of the most complete maps of a plant’s genetics. By identifying genes with great precision, it will be useful for breeding plant species other than the rose, as well, he said.
Now, to develop a new type of rose, breeders typically make thousands of hybrid offspring, looking for the combination of traits they want. Then, they have to select and identify the offspring that have the desirable trait. It’s a process that can take up to 10 years and require lots of greenhouse space and land, as well as water, said Mohammed Bendahmane, a senior author on the paper and research director at the École Normale Supérieure de Lyon, in France.
With data from the more detailed sequence of the rose genome, this process should be significantly shortened, reducing the cost and energy consumption needed to introduce new species, he said.
Because of centuries of breeding, most of the modern rose cultivars have four copies of genes, two from each parent — rather than the more typical one from each parent. This complexity makes the genome tricky to sequence and to assemble. To circumvent this, the researchers created a rose with just a single copy of each of the genes.
Dr. Bendahmane and his colleagues and partners started with a rose variety called Rosa chinensis “Old Blush,” which originated in China and was introduced to Europe in the 18th century. European rose breeders hybridized their plants with some from China to take advantage of the continuous blooming, scent signatures and color of the Asian plants.
The researchers also sequenced genomes from ancestral rose species and newer hybrids to understand the composition and the structure of modern roses and the origin of important traits.
“Now we can combine the information from genetics that have been done before, together with our data from the genome, including gene diversity and structure, to discover which of the ancestral botanical roses participate in which trait,” Dr. Bendahmane said.
Up-to-date gene sequencing technology also allowed the team to develop a more detailed genetic map, said Todd Mockler, a principal investigator at the Donald Danforth Plant Science Center in St. Louis, who was not involved in the new research.
“If you only have 80 percent of the genome, you wonder what’s in the 20 percent you’re missing,” he said, noting that previous sequences often missed genes involved in disease. “The completeness is a big deal.”
Dr. Mockler, whose team sequenced 400 plant genomes last year, said the paper marked a new “democratization” of plant research. A decade ago, a study like this one would have cost $20 million or more, he said, and would have been feasible only for high-value, high production crops like wheat, corn and soy. Now, he said, such detailed sequencing is becoming much cheaper and more widely available.
Editing the genes of crops like roses — to reduce pesticide and water use, for instance — will also become more realistic now that there’s a good road map of those genes, he said.
“The big challenge is you need to know what to edit,” Dr. Mockler said. “You can’t just randomly start editing. You have to know what to target. The only way to know that is to have a genome sequence.”